How can retailers turn AI agents into real profit drivers?

These tools let retailers adjust prices and refine promotions in near real time.

Retailers should rethink how their merchandising teams work as so-called agentic artificial intelligence (AI) systems move from experimentation into day-to-day operations.

Unlike traditional analytic tools, agentic AI is built to plan, act and learn on its own. In merchandising, that could let retailers adjust prices, fine-tune promotions and rebalance product assortments in near real time, replacing the static weekly or monthly review cycles that still dominate many organisations, according to consulting firm McKinsey & Co.

“Historically, the merchant role requires spending critical time and energy on manual, repetitive tasks,” it said in a January report.

“By offloading it to AI agents, merchants could reclaim up to 40% of their time to do what they do best: focus on strategy, find great products, understand customers, and optimise vendor negotiations,” it added.

Yet most retailers are not ready to deploy the technology at scale.

McKinsey found that 71% of merchants said AI merchandising tools have had limited or no impact on their business so far, whilst 61% said their organisations were only slightly prepared, or not prepared at all, to scale AI across merchandising functions.

One barrier is data quality. Generative AI tools often struggle with fragmented commercial data, such as inconsistent stock-keeping unit files or incomplete pricing histories.

That results in insights that are “directional rather than decision-ready,” according to the report.

Agentic AI systems are designed to address those gaps. They can clean and reconcile data autonomously, run thousands of scenarios at once and coordinate actions across pricing, promotions, inventory, and vendor management.

Crucially, they are also built to explain the reasoning behind their recommendations, a key requirement for merchants who remain accountable for results.

“Early agentic AI adopters are already seeing the benefits of this new frontier, with significant revenue and margin lifts resulting from stronger assortment decisions and data-backed bargaining capabilities,” McKinsey said.

The market opportunity is drawing attention. The agentic AI segment in retail and e-commerce is projected to reach about $60b this year, up from $47b a year earlier, and could exceed $218b by 2031, according to data from Mordor Intelligence Private Ltd.

Nvidia Corp.’s third annual State of AI in Retail and Consumer Packaged Goods survey found that retailers are using AI agents to power digital shopping assistants, dynamic product catalogues and faster warehouse and supply-chain operations.

In the survey, 47% of retail and consumer goods companies said they are using or evaluating agentic AI, whilst 21% plan deployments within a year.

McKinsey said the technology could reshape daily work for category managers. Instead of reacting to problems after the fact, merchants could start each day with AI-generated briefs flagging underperforming promotions, local pricing gaps or emerging markdown risks.

With limited human approval, agents could then adjust store-level prices, redirect marketing budgets or shift inventory.

Vendor discussions could also change, moving away from backward-looking reviews toward forward-looking growth planning, supported by live benchmarks and scenario analysis.

Still, McKinsey warned that tech alone would not deliver results. Retailers need to redesign operating models, clarify decision rights and invest in skills.

Only 24% of merchants surveyed said they receive substantial AI upskilling.

The firm also flagged risks, including unclear pricing guardrails, overreliance on short-term metrics, and resistance to moving away from legacy systems.

To manage those risks, McKinsey urged retailers to roll out agentic AI in phases, pairing technology with workflow changes and scaling only after consistent performance is proven.

Questions to ponder:

1. How should retailers redesign merchandising roles as AI agents take on more analytical work?
2. What safeguards ensure autonomous pricing and promotions protect brand trust and profits?